Licensed under the Creative Commons attribution-noncommercial license, http://creativecommons.org/licenses/by-nc/3.0/. Please share and remix noncommercially, mentioning its origin.

Produced in R version 3.2.4 using pomp version 1.4.7.1.
Objectives
- display a published case study using plug-and-play methods with non-trivial model complexities
- demonstrate the use of profile likelihood in scientific inference
- discuss the interpretation of parameter estimates
- emphasize the need to allow for extra sources of stochasticity in modeling
Measles revisited
Motivation: challenges in inference from disease dynamics
- Understanding, forecasting, managing epidemiological systems increasingly depends on models
- Dynamic models can be used to test causal hypotheses
- Real epidemiological systems:
- are nonlinear
- are stochastic
- are nonstationary
- evolve in continuous time
- have hidden variables
- can be measured only with (large) error
- Dynamics of infectious disease outbreaks illustrate this well

Outline
- revisit classic measles data set
- ask questions about:
- measles extinction and recolonization
- transmission rates
- seasonality
- resupply of susceptibles
- use a model that
- expresses our current understanding of measles dynamics
- cannot be fit by existing likelihood-based methods
- examine data from large and small towns using the same model
- does our perspective on this disease change?
- what bigger lessons can we learn regarding inference for dynamical systems?
He, Ionides, & King, J. R. Soc. Interface (2010)
Data sets
- Twenty towns
- population sizes: 2k–3.4M
- Weekly case reports, 1950–1963
- Annual birth records and population sizes, 1944–1963

